Serving up the PIE: The Importance of the Right Withdrawal Strategy

We have written about our investment portfolio more than once and the purpose of this post is not to go over that again. What we are going to get into is how we (a) plan to position our portfolio for early retirement and (b) our withdrawal strategy once we get there.

There are many great articles out there on each of these two topics. Indeed, probably more articles than there are actual mutual funds available – and that is a lot. The Physician on Fire has described his Investor Policy Statement, the Bogleheads forum must have thousands of articles on asset allocation strategies and Wade Pfau has documented in many peer reviewed articles the dizzying array of options available to the early retiree on withdrawal strategies. To be honest, the amount of information out there can be overwhelming. Here is a “Simple Simon met a PIEman” approach that tries to distill it all down to something that we and you can understand. Or as Einstein famously said:

Everything should be made as simple as possible, but not simpler.

First of all, here is a summary of our plans.


  • Retire early (Mr. PIE age 51; Mrs. PIE age 45) with two boys in elementary school (age 11 and 9).
  • Target date is July 3, 2018.
  • Relocate the family to our mountain home in Northern New England.

Asset Allocation at Early Retirement:

  • Portfolio target with approximately 50:50 split between taxable and tax-deferred investments.
  • Portfolio mix to be 65% equities, 25% bonds, 5% REIT, 5% cash.
  • US to ex-US equities allocation  ~5:1 with international equities spread across total international, developed and emerging markets.

Early Retirement Income Sources:

  • Company pension (Mr. PIE) starts paying out at early retirement as “income floor” that will cover ~35-40% of expected annual expenses in retirement.
  • The dividends from our taxable account will be directed to our bank checking account on a regular basis. Current asset allocation mix in our taxable account has a weighted dividend yield of 2.65%.
  • Sell taxable assets from equities or bond bucket according to CAPE (Cyclically Adjusted Price to Earnings) ratio or Equal Withdrawal for each. We’ll get into the meat of this later.
  • Mrs. PIE may elect to bake and decorate a spectacular cake or two, sell them and direct any side hustle profits to family needs such as kids college funds or Mr. PIE Sunday afternoon adult beverage ventures…

Other Factors:

  • Conservative investing approach, which has actually moved from ultra-conservative to just boringly conservative. Sleeping extremely well is the name of our game. It is as simple as that.
  • 529 accounts are not factored into our asset allocation or withdrawal plan. Those funds are already spoken for.
  • Taxable account is with Vanguard in low cost passive index funds.
  • Our employer sponsored 401K’s will each be rolled over into Vanguard at separation from service and assigned according to the equities/bonds/REIT allocation described above.
  • We don’t intend to invoke any 72t distribution or Roth conversion strategy with our tax-deferred investments. We will however assess RMD’s (Required Minimum Distribution) projections well in advance of them kicking in at age 70.5.

Annual Expenses

  • Minimized through (a) zero mortgage  (b) travel hacking to fund majority of vacation budget and (c) living in a location within a tax-friendly state – i.e. low property taxes and no state taxation of pension income. Our tax obligation is projected to be very low according to running the numbers through


The PIE Withdrawal Strategy

Financial planning gurus like Wade Pfau and Todd Tressider have written and spoken about different withdrawal strategies. Go listen to this podcast and this one for great discussions between Wade and Todd on different withdrawal strategies to support income generation.

It was also very timely that Fritz over at the Retirement Manifesto outlined a common approach known as the Bucket Strategy that he is putting in place. On that very post, Mr. PIE commented on the approach of family PIE as being roughly similar to his. Indeed we also hold some of the same index funds within the different buckets. Please go take a look at his post that featured on RockStar Finance (Congrats Fritz!) last week. Anyway, not to recreate what Fritz was describing, here is a summary of “our buckets” with some nuances that are unique to us and perhaps others.

Bucket 1

This is the cash bucket that we described above as 5% of our portfolio. Our existing emergency cash on hand will be increased using a portion of the proceeds from the sale of our primary residence in 2018. The number of years of expenses depends on your risk tolerance and factors such as other base income sources (e.g. pension). As we described above, we will have a pension thus our cash on hand has that factored in and we will have a lower amount in this bucket than is typical for an early retiree with no pension income. For the well documented Sequence of Returns risk, this is why holding cash is king to avoid selling equities or bonds in a prolonged bear market.

Bucket 2, Bucket 3

Here is where it gets tricky.

Where are you going with this, Mr. PIE?

Is there a hole in your bucket(s)? Are they going to empty too fast?

Hopefully no holes, just some reasonable arguments perhaps. Let’s go.

We described above that our asset allocation with be a mix of safer asset classes like bonds (typically Bucket 2 in the traditional bucket approach) and riskier asset classes like equities (typically Bucket 3). Beyond our income floor pension (barring a scandal of Enron proportion and Mr. PIE’s company going out of business, we expect to continue to receive this in retirement!) and regular payout of dividends from our taxable account, we will need to sell some assets to meet our annual expenses.

The important question to understand is – which asset class should we sell on an annual basis?

Again, this is where some fine withdrawal strategy research is out there on the interwebs. Darrow Kirkpatrick who blogs at Can I Retire Yet did some truly excellent work on this very topic. To summarize, Darrow used historical data to test a portfolio of equities and bonds against different withdrawal strategies. In his original article, he tested six different methods based on a 50:50 mix of equities : bonds using a $1M portfolio and 4% withdrawal rate. He then computed the results for each 30-year retirement span from 1928.  There were surprising results across each of the strategies. The table below highlights four of the strategies and the ending portfolio value after 30 years.


In some subsequent work, Darrow added in different asset allocation portfolios (60/40 and 80/20) and also incorporated an annual rebalancing strategy into each method. The bottom line is no different – CAPE strategy still wins. There is no CAPE Fear.

So what is CAPE (Cyclically Adjusted Price to Earnings)?

CAPE is a valuation measure usually applied to the S&P 500 and is defined as price divided by the last ten years of earnings, adjusted for inflation. It is conveniently available with the click of a mouse and heading over to this web-site.  The current Shiller PE ratio is 26.5 compared to the median of 16. Higher values typically indicate stocks may be overvalued. What it is telling the early retiree to do according to the CAPE withdrawal strategy is to sell equities at the end of 2016, assuming the ratio does not move significantly. In practical terms, if you have a $1M 50:50 equities:bond portfolio with $40K in expenses and need $20K to add onto the $20K of dividends that a $1M portfolio throws off, simply sell $20K worth of equities. Simple as that.

And if you don’t believe what Darrow has to say about using CAPE as part of a long term withdrawal strategy, go read this article from another great financial planning researcher Michael Kitces on other ways CAPE can help in the context of your retirement plans.

The data in the table above provides a few simple conclusions:

  1. All four strategies provide a sizeable ending portfolio after 30 years.
  2. Yet there are startling differences in ending portfolio value. In fact >$4M from just the four different methods described above!! That is some serious moola.
  3. Adopting the CAPE strategy appears to afford a significantly larger portfolio through retirement and ultimately to pass on as legacy to children, extended family and/or charity. Who wouldn’t want either of those scenarios?
  4. The equal withdrawal method is certainly the simplest if you are too lazy to click a mouse and go over to the Shiller site. Any retiree should have ample time to fire up their lap-top or mobile device and find the current PE ratio. Simple as that.

Market headwinds will change over time. Past performance is no guarantee of future performance. This is stating the obvious. You know that flexibility in your spend is the adaptability factor that needs to be invoked depending upon the behavior of the market. And ultimately how you manage the trajectory and size of your portfolio.

Unless a better looking strategy materializes in the retirement planning literature, the CAPE method seems like a very powerful way to go. Or do you see any major impediment to adopting this strategy?

Of course, nothing is ever simpler. Perhaps just as simple as possible.

For those in early retirement or approaching it, what are you doing in terms of a withdrawal strategy? What do you make of the research and findings behind the CAPE withdrawal strategy?


  1. The bucket strategy is very intuitive and I’m a fan. I actually saved a Morningstar article intending to go back and reference someday. I essentially have “mini buckets” for the mini retirement / long road trip.

    I wasn’t familiar with median CAPE. That is also intuitive.

    1. Thanks TJ. I think there is a Morningstar video with Christine Benz dedicated to the whole bucket strategy concept. And good luck with your buckets on the long road trip. Pack them safely!

  2. The CAPE strategy is a clear winner. Very interesting. I will keep that in mind. I hope not to have to think too much about selling I’ll have a 457(b) that will be paying out like a pension, and about a 2% dividend from the index funds in the taxable account. The combination won’t be enough; I may have to beef up the blog income between now and Then.

    Oh, and Then is looking like a date very close to yours. Tentatively looking at June 30, 2018 as the date I leave my current job. I’ll plan on doing some locums in NZ or AUS because I can, but after that, I believe I’ll be ready to call it a day.

    Thank you kindly for mentioning my IPS. It’s a handy little document.


    1. We found the IPS concept very useful to formulate our plans in our mind. So much easier when it is down in writing and you see the interplay of various things. There is certainly much to think about in regards to withdrawal strategy. The more I read, the less simple it gets. I guess being consistent is a correct answer, whatever strategy one decides to adopt.

      Summer of 2018 is for sure going to be exciting times for our respective families and I am sure your research into AUS or NZ is going to be fun digging into and writing about.

  3. Very interesting Mr PIE. That does make sense to me, why CAPE would be a good indicator to sell. Excessive bull markets tend to overvalue stocks (more than they’re worth, good to sell) and in a recession/crash they undervalue them, bad to sell.

    The one ‘danger’ I suppose is what is your average CAPE (and time period) that you’re comparing to? Does the low interest rates justify some values or not? I suppose it’s hard to say. But the premise of overvalued and undervalued stocks in bull and bear markets would still work.

    I read quite a few Australia financial (not necessarily pf blogs), and there’s one in-particular that I like. I remembered one article in the last couple of months about CAPE, so I thought I’d link it to you. It isn’t supporting or opposing your idea, or Darrow’s, I think it’s actually agrees with your conclusion that being significantly above CAPE is a good time to sell stocks.

    Anyway, here it is:


    1. Thanks for the link to the article Tristan and the useful addition to the discussion. The work of Darrow has the CAPE number taken from the long term median. Interest rates are for sure a complicating factor as Mr. FL also indicates below in the comments. Who knew how to withdraw your hard earned investments would be so darn hard?!

  4. I like the analysis, Mr. Pie. I’ve only taken cursory looks at withdrawal strategies, but the CAPE seems logical. It feels like timing the market, which I’m not a fan of, but it has a pretty proven track record.

    Personally, I’m not sure what my stock:bond mix will be in retirement. I’m leaning towards a heavy stock mix (perhaps all stock if I’m comfortable enough with the size of my nest egg) which may make the CAPE strategy irrelevant. We’ll see though. Thanks for sharing this!

    1. Yes, indeed. Having an all equities portfolio is one way to avoid the decision making on which asset class to sell!! Although I don’t see too many retirees of age 50+ doing it. Especially those who don’t want to go back to work if the portfolio takes a brutal hammering in a bad down market. It can happen as history teaches us.

      Thanks for the perspective.

      1. As a fellow “near FIRE” blogger, I agree with Mr PIE that few of us are anywhere near 100% equity. Yes, the highest long term return, but not worth the sequence risk as you near retirement. I’ve been an aggressive investor my whole life, but am now 50:50 as I close in on 18 months to retirement.

  5. Very nice research! It seems that a lot of people are thinking about this topic now! It’s on our mind as well, 18 months away from retirement.
    I did some research and found that the CAEY (cyclically-adjusted earnings yield = 1/CAPE) is a pretty good predictor of good withdrawal rates. For a 100% equity portfolio, the 1%+0.5*CAEY (currently 0.01+1/26.5 = less than 3%, ouch!) rule worked pretty well, much better than the 4% static.
    Bonds are the great unknown: Mixing in too much of a currently low-yielding asset and your expected returns will not sustain your withdrawal rate. And all that for not really that much diversification benefit.
    Also: Personally, I would never look at median final portfolio values. Median values are, by definition, astronomically high because the median real return easily beats the withdrawal rate (at least if you have enough equity share). I’d be more interested a) a measure of risk/standard deviation, especially the downside risk (e.g. what’s the lowest and 5th-percentile lowest value final value) and b) a measure of the variance and minimum value along the way, not just at the end point.
    Just a thought!

    1. Thanks ERN for the very thoughtful response. Just reading in preparation for this post offered a dizzying amount of strategies. I see a lengthy trail on the Bogleheads forum on CAPE withdrawal and other strategies including variable percentage withdrawal (VPW). I also noticed on another forum the discussion point you raised on volatility along the way. Darrow did describe the success rates in both articles and I am sure he has the 5th percentile data. I imagine he may put out some more work on this topic as it was one of his most popular posts as he described himself on his blog.

  6. First off, this is a totally killer post. Truly awesome work! I love the detail and logic on display, and y’all are kicking serious backside when it comes to planning.

    Now, the idea to use the CAPE-10 to determine where/how to make withdrawals makes some theoretical sense. After all, median CAPE is the best long-term measure of central tendency when it comes to equity valuation. But it’s also a horrible short- and mid-term indicator of valuation…as in, equities can be “wrongly” valued for incredibly long stretches. In other words it’s not a super-awesome planning tool.

    “But, FL,” you say, “it way out-performs those other withdrawal strategies.” Yes, but that’s probably only because those other approaches are really, truly terrible. Withdrawals on the basis of last year’s performance or the last seven years’ performance are effectively arbitrary strategies that don’t account for relative valuation. So I’d say those are worse than shots in the dark since they’d tend to cause selling of equities simply because equities are a better long-run investment than bonds. And the equal withdrawals approach generates a good result only because it benefits from the absence of trying to time the market in some way like recent performance timing does.

    So CAPE isn’t terrible. But CAPE is kind of troubling, too, since there are some complications in measuring it, it doesn’t seem to reconcile interest rate issues very well, and a CAPE drawdown program ignores bond valuation issues.

    An alternative regime that might be worth thinking about that takes direct account of interest rates and, by extension, debt valuation, is looking at historical equity valuation on the basis of market-wide “EBITDA / Total Enterprise Value.” This also involves some measurement complexity. But it can arguably more cleanly account for firms’ capitalization decisions and GAAP-related accounting distortions than CAPE. It has the added benefit of linking equity and debt valuation. One way to consider using this measure would be to compare the current or latest reading against a long-term central value and to overweight equity sales when the measure is well above the central value. Just a thought. A little more complicated than CAPE but theoretically more rationalized – in my opinion. This is a complex area, after all, and, because of the long-term horizons being dealt with, any strategy might perform strongly (or poorly) despite solid theoretical grounding over an individual’s retirement phase!

    1. Thanks for the vote of appreciation. I do feel like the summer intern writing about stuff to the tenured professors! The great mix of our blogging community I guess!

      Regarding the EBITDA / Total Enterprise Value, where would one find such information / data to read and learn more? Where do such latest numbers reside?

      And completely agree about the unique nature of our situations – age, horizon (insert predicting death), portfolio size, expenses….Adjusting spend and being adaptable is a constant theme no matter which particular withdrawal strategy is being discussed on forums such as Bogleheads, MMM etc.

      1. Don’t want to oversell this EBITDA multiple thing because it also has shortcomings. But it can provide a different perspective and/or augment other considerations like CAPE.

        Damodaran over at NYU periodically compiles and publishes EBITDA/EV data in Excel-ready tables. For historical comparisons, Bloomberg data are probably the easiest to deal with, but may be subscription only. Your brokerage account research tools likely allow for screening on EBITDA/EV or its reciprocal (that’s what most traders use this multiple for, to compare equity investments), so you can probably run it against indexes to see where the market resides at a given time and get histories, etc. Long-run avg. EV/EBITDA is around 11.0 for the S&P500, and we’re around 13 now, which is a pretty big departure from “normal.” (You’ll see that on a market-wide basis this metric doesn’t tend to fluctuate as much as some other statistics.)

        Again, don’t want to overstate the importance of this metric. My fondness for including it in broad considerations is that it pulls capitalization mix into consideration…which, theoretically, has some merit as a forward-looking indicator since over-leveraged firms that face declining relative earnings will get brutalized in market cap, and when there’s over-leveraging in the entire market…well. So it’s part of a group of things that can make sense to look at when making withdrawal decisions that attempt to beat the averages.

        Hope this helps. We should move this discussion over to the bar, I think!

        1. Thanks for adding to the discussion and my poor head now hurts. Yes, let’s make it hurt more after a few drinks in a fine bar.

          Are you going to FinCon next year, by the way?

  7. I think it’s a great plan! Only issue I seeing is you having to do something about the tax deferred accounts sooner than later due to risk of having high RMDs pushing you into higher tax brackets.

    Haven’t read about the CAPE methodology yet. I always assumed I’d do the withdrawals based on asset allocation. I also plan on retiring in my 30s with no intentions of not earning income again, so my plan is very fluid right now (or lack there of an actual withdrawal strategy at this point).

    1. FF, thanks for swinging by. The RMD problem is indeed a first world problem for some! Although as even the IRS tell us, we have to take them but don’t have to spend them. Death and some level of taxation are unavoidable.
      I can fully understand you not thinking about this stuff yet. Still, with all the information out there on this topic, best to start reading sooner than later. The material is actually mind-boggling when you start digging…

  8. I’ll admit that this is the most I’ve read on the CAPE methodology to date so I’ll have to follow your links and study up. I’ve always been a dollar-cost averaging guy at about 90% stock allocation. Of course that will change when we hit our FIRE date.

    Also, I definitely plan to climb the Roth IRA Conversion ladder! We have most of our retirement income tied up in 401k/IRA so getting it out to live on will be key after the first 5 years.

    Thanks for all the research and info, it’s great!

    1. Thanks for checking us out HNRE!

      Definitely take in the great work of Darrow. There’s also a good discussion on the Bogleheads forum that provides support for CAPE withdrawal and some counter arguments. Worth a look. It gets very feisty some of the discussion…

      Your Roth conversion strategy makes sense. With that one, there is a clear roadmap of what to do and how to do it.

  9. Wow, Mr Pie! I can’t tell you how honored I was when I read this post and saw the incredibly kind words about my recent article. I LOVE that you wrote this blog and built on the concept with CAPE! After you left the comment on my blog re: CAPE (which I acknowledged favorably), I actually came across a note in my withdrawal strategy tab in Evernote which I had written 6 months ago to investigate CAPE for potential inclusion in my strategy to help determine when to “refill buckets”. We think AMAZINGLY alike, and I’m very pleased that you built on the bucket strategy, and added valuable content to the discussion. Thanks for sharing the credit! Keep up the good work!

    1. Hey Fritz,

      Thanks for the kind words.

      Sometimes it takes others to provide validation of our own thinking.

      The real credit lies with Darrow and his two excellent posts on the topic. It is interesting that there is not an awful lot out there on which asset class to sell and why. Darrow has really brought this topic to the fore.

      It’s not easy this FIRE stuff and I’m glad there are so many tremendous resources out there for us all to learn from!

  10. Ok, I’ll throw in another curve. I had lunch with my accountant last week (buy then lunch NOW, when they’re not busy, and talk strategy). I’m also planning on minimizing long term withdrawals from pre-tax accounts by “topping off” my income to the next income tax bracket. If I’m $5k from the next tax rate, I’I’ll withdrawal $5k from my IRA (yes, in my late 50’s, far ahead of any RMD’d). This allows you to gradually reduce RMD’s, and pay some tax on your IRA money at your lowest possible tax rate. I’all likely do it for 15 years (from Age 55 to 70 1/2). I’ll use the CAPE ratio to decide whether to sell equities or bonds from my IRA. Anyone else looking at this?

  11. We’ve recently begun looking into withdrawal strategies as well, because I’m at least looking at Summer 2018 as a no more work date, even if Mrs. SSC continues teaching. We’re still high equities vs bonds though – something else we’re working on how best to protect the nest egg while still getting decent gains.

    I actually signed up for a free financial advisor session thru work and I think I blew his mind. 🙂 I gave him the 5 minute version of where we are financially and our plan and then we discussed different income generating withdrawal strategies the remainder of the session. It was actually a really good time and very informative.

    Like you said, the info on this is mindboggling and can send you down so many rabbit trails. I hadn’t looked at using CAPE or EBITDA/EV as indicators, so this post is timely. Thanks for the links and discussion in the comments!

    1. Glad you could have a useful session with your client…! :>)

      The equities vs bonds ratio is giving us some headaches also. My pension income stream can afford us to go a bit more aggressive but we just can’t do a very heavy equity ratio.

      I suspect there will be much more to come out on withdrawal strategies to dizzy the mind. There was an interesting article just this week over at Morningstar by Rekenthaler on the failure of the 4% rule and provided new insights from a global perspective. Worth a read:

  12. This is awesome! I’m quite a ways from hitting our financial independence number so I hadn’t yet been looking as much at withdrawal strategies, but this has gotten my brain churning.

    I wonder if anyone has compared how using the different methods could allow you to have a higher overall safe withdrawal rate. The example above normalizes on withdrawal numbers and compares portfolio values. If you could solve for the other side – normalizing on portfolio values, you could solve for the withdrawal rate and potentially use a CAPE strategy to retire at the same “risk level” as other strategies but with a smaller balance.

    Thanks for putting this together and getting me thinking about how to manage the other phase of FI!

Leave a Reply

Your email address will not be published. Required fields are marked *